Cycle stability is important for preparative chromatography resins. Up to 200 cycles have been reported for Protein A affinity resins when used under optimized operating conditions. Through engineered ligands, alkaline resistant Protein A resins are available that can withstand repeated cleaning-in-place cycles with even 1 M NaOH.
Journal of Chromatography A 1673 (2022) 463058 Contents lists available at ScienceDirect Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma Alkaline treatment enhances mass transfer in Protein A affinity chromatography Nico Lingg a, Andreas Daxbacher a, Desiree Womser-Matlschweiger a, Dietmar Pum b, Jürgen Beck a, Rainer Hahn a,∗ a Department of Biotechnology, Institute of Bioprocess Science and Engineering, University of Natural Resources and Life Sciences, Vienna, Muthgasse 18, Vienna 1190, Austria Department of Nanobiotechnology, Institute of Biophysics, University of Natural Resources and Life Sciences, Vienna, Muthgasse 11, Vienna 1190, Austria b a r t i c l e i n f o Article history: Received 28 February 2022 Revised 11 April 2022 Accepted 11 April 2022 Available online 14 April 2022 Keywords: Mass transfer Preparative chromatography Antibody purification Pore diffusion Solid diffusion a b s t r a c t Cycle stability is important for preparative chromatography resins Up to 200 cycles have been reported for Protein A affinity resins when used under optimized operating conditions Through engineered ligands, alkaline resistant Protein A resins are available that can withstand repeated cleaning-in-place cycles with even M NaOH This enables an increase of purification cycles through the reduction of fouling while maintaining high binding capacities Previously, non-intuitive changes in dynamic binding capacity after alkaline treatment have been observed for these novel Protein A resins, where sharper breakthrough curves and increased capacities were reported In this work, we have systematically investigated resins with both low and high alkaline stability and studied the changes in static and dynamic binding capacities and elution behavior We propose that the observed mass transfer increases of up to 40% are due to a switch in diffusion mechanism, as shown by confocal laser scanning microscopy Based on our results, only a small window of alkaline treatment conditions exists, where dynamic binding capacity can be increased Our findings may help to explain previous findings and observations of others © 2022 The Authors Published by Elsevier B.V This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/) Introduction Protein A affinity chromatography is the backbone of platform processes for the purification of monoclonal antibodies (mAb) from cell culture supernatants One reason for its popularity is the simplicity of the purification scheme: binding at neutral pH and almost completely independent of salt concentration followed by elution at acidic conditions Hence, method development is straight forward and easy, which has contributed to the almost universal application of Protein A affinity chromatography at the core of antibody platform purification processes Since the introduction of the first commercial Protein A resin in 1978, tremendous efforts to improve chromatographic performance were undertaken and are still continuing today, more than 40 years later [1] Dynamic binding capacities (DBC) of early resins were in the range of 5–10 g IgG per L and have been increased approximately ten-fold since [1] Modern Protein A affinity resins have capacities similar to other commonly used resins, such as those for ion exchange and hydrophobic interaction Numerous comparative studies have ∗ Corresponding author E-mail address: rainer.hahn@boku.ac.at (R Hahn) been performed and published that document this improvement over the years [1–9] This vast capacity increase has been achieved by several strategies Recombinant Protein A ligands consisting of up to domains have been designed and attached to the resins at ever increasing ligand densities Moreover, mass transfer resistance was reduced by increasing the pore size, which has led to improved utilization of the static binding capacity Steadily increasing IgG production titers in cell culture and a general drive towards improved process economics have generated a demand of Protein A resins that can be used over many purification cycles [10] Together with increased regulatory requirements, this has led to a demand for efficient cleaning procedures to minimize resin fouling and to prevent carry-over of impurities The stability of Protein A ligand against sodium hydroxide, the most popular reagent for sanitizing and cleaning in place (CIP), has been increased substantially This was achieved through genetic engineering tools, e.g substitution of asparagine and glutamine residues that are responsible for the structural destabilization of the Protein A domains upon alkaline treatment [11–15] These improvements have led to stationary phases that are stable for hundreds of CIP cycles with 0.5–1.0 M NaOH However, a very small reduction of binding capacity is observed after each sanitization cycle Typically, such data is provided https://doi.org/10.1016/j.chroma.2022.463058 0021-9673/© 2022 The Authors Published by Elsevier B.V This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/) N Lingg, A Daxbacher, D Womser-Matlschweiger et al Journal of Chromatography A 1673 (2022) 463058 in the data sheet by the manufacturers obtained from long-term cycling studies Several publications deal with the deterioration of Protein A resins throughout their life-cycle Pathak and Rathore investigated the influence of up to 100 purification cycles on MabSelect SuRe with cell culture supernatant and 50 mM NaOH used for CIP [16] After 100 cycles, the resin exhibited an earlier and shallower breakthrough curve, resulting in reduced dynamic binding capacity and mass transfer They observed a physical film of foulant by transmission electron microscopy (TEM), which led to pore constriction Zhang et al used MabSelect resin in process scale with crude material, which was regenerated with 50 mM NaOH for 20 cycles [17] A constriction of the pores due to fouling was observed, which led to a reduction in mass transfer and dynamic binding capacity [18] This decreased mass transfer was confirmed by confocal laser scanning microscopy (CLSM) These drastic fouling effects can be attributed to the low caustic concentrations used for cleaning Hahn et al have shown that fouling effects can be highly dependent on the combination of resin, feed material and cleaning protocol applied [7] Yang et al investigated strategies to protect MabSelect and MabSelect SuRe ligands from alkaline degradation [19] They observed a reduction in affinity of the MabSelect ligand after alkaline treatment with 20 to 200 mM NaOH This decrease was mostly driven by an increase in the dissociation rate They also observed a reduction in static and dynamic binding capacities, dependent on NaOH concentration after up to 100 purification cycles with purified IgG The breakthrough curves also show a change in slope, with new resin having a shallower breakthrough than alkaline treated resin Jiang et al investigated the influence of alkaline CIP cycles on rProtein A Sepharose Fast Flow and MabSelect resins using pure IgG [20] These older resins are not alkaline stable and as such 50–100 mM NaOH was used for cleaning Adsorption isotherms and breakthrough curves showed a reduction in static and dynamic binding capacity, and a change in the slope of IgG breakthrough was observed The cycled rProtein A Sepharose FF resin had steeper breakthrough curves than the new material The authors hypothesized that this changed slope originated from altered mass transfer properties and suggested that further investigations are required to unravel the origin of this phenomenon While fouling clearly leads to decreased mass transfer due to pore constriction, the effect of alkaline treatment alone appears to be more complex The effects of such CIP treatment on the intrinsic mass transfer mechanism has not been investigated so far While CLSM studies exist for cycled Protein A resins [17,18], these studies did not investigate the effect of alkaline treatment alone The purpose of this study was to investigate potential changes of mass transfer characteristics upon alkaline regeneration with varying NaOH concentrations, in line with the suggestions for further investigations as proposed by Jiang et al Additionally, it was intended to reveal differences between the highly alkaline stable Protein A resin MabSelect PrismA and its predecessor MabSelect SuRe To broaden the scope of investigation, a methacrylate-based resin, Toyopearl AF-rProtein A HC-650F, was also studied The lig- ands used in MabSelect and Toyopearl Protein A resins are derived from different domains [15,21,22] We have used experimental CLSM data and compared them to theoretical CLSM images to uncover the mass transfer mechanisms [23] Experimental 2.1 Materials and instrumentation MabSelect SuRe and MabSelect PrismA were purchased from Cytiva (Uppsala, Sweden) and Toyopearl AF-rProtein A HC-650F from Tosoh Bioscience (Stuttgart, Germany) The properties of these resins can be found in Table In this study, two different IgG samples were used: polyclonal IgG was a kind gift from Octapharma (Vienna, Austria) A monoclonal antibody, an Adalimumab biosimilar, was produced in-house using recombinant CHO cells The antibody was purified by Protein A affinity chromatography and the eluate was buffer-exchanged on a Sephadex G25 column (Cytiva) The activated fluorescent dyes for protein labeling, Rhodamine RedTM-X Succinimidyl Ester and Pacific BlueTM Succinimidyl Ester were obtained from ThermoFisher Scientific, Life Technologies Corporation (Grand Island, NY, USA) Buffer chemicals were purchased from Merck (Darmstadt, Germany) and SigmaAldrich (St Louis, MO, USA) if not stated otherwise Chromatographic runs were performed on an ÄKTA Explorer 100 using resins packed in Tricorn 5/50 and 5/100 columns (all Cytiva) Confocal laser scanning microscopy was performed using a Zeiss LSM 510 microscope with a Plan-Apochromat 63x/1.4NA oil objective (Carl Zeiss MicroImaging, LLC, Thornwood, New York, NY, USA) and a Leica Sp5 microscope with a Plan-Apochromat 40x/0.85 dry objective (Leica Microsystems, Wetzlar, Germany) 2.2 Methods 2.2.1 Alkaline treatment Resin packed in columns were treated with NaOH at varying concentrations and incubation times as stated in the results section For adsorption isotherms and CLSM measurements, the incubation was instead performed in batch After the end of the stated incubation period, the pH was neutralized using an extensive PBS wash 2.2.2 Breakthrough curves and linear pH gradient elution For the determination of the dynamic binding capacity, breakthrough curves (BTC) were performed at varying velocities ranging from 50 to 600 cm/h The BTCs were performed in Tricorn 5/100 columns with g/L polyclonal IgG Two evaluations were used: firstly, the breakthrough profiles were fitted by a constant pattern column adsorption model with film and pore diffusion [24] and secondly, a rearranged model for pore diffusion was used to derive De from DBC10% data The equations for this model can be found in the supporting information After loading the column, bound IgG was stripped using 100 mM glycine, pH 2.5 Table Properties of Protein A resins from Pabst et al [5] DBC at 2.4 residence time Parameter Matrix MSS agarose PrismA agarose 650F polymethacrylate Dp (μm) rpore (nm) ε p (NaCl) ε p (mAb) EBC (mg/mL) DBC (mg/mL) De (10−8 cm²/s) Protein A ligand progenitor domain (number of repeat units) 89.6 41.8 0.96 0.76 50.2 34.5 4.8 B (4) 54.1 32.7 0.94 0.50 85.4 49.0 1.1 B (6) 54.2 30.8 0.91 0.36 61.5 42.4 1.6 C (6) N Lingg, A Daxbacher, D Womser-Matlschweiger et al Journal of Chromatography A 1673 (2022) 463058 Fig BTCs with IgG on MabSelect SuRe (left panel), MabSelect PrismA (middle panel), Toyopearl AF-rProtein A HC-650F (right panel) Column height was ∼10 cm, velocities were 75, 150, 200, 300 and 600 cm/h Table De and qm values determined from different model fits for the investigated Protein A resins For the study of time and concentration dependent effect of alkaline treatment on DBC10% , a Tricorn 5/50 was used instead After reaching the incubation time the column was equilibrated and a BTC was performed After the BTC experiment a linear pH gradient elution run was performed The same column was subsequently further incubated to reach the following time of incubation stated in the results section, i.e after an initial 12 h incubation the column was incubated for another 12 h to determine the 24 h time point The pH gradient was achieved by using a linear gradient of 50 mM citrate buffer from pH 5.5 to 2.5 over 15 column volumes The column was loaded with mg IgG per mL packed bed Experiment/model fit Parameter MSS PrismA 650F DBC vs tR global fit De qm De qm De qm 5.9 × 10−8 47 5.6 × 10−8 58.5 8.0 × 10−8 28.9 2.0 × 10−8 80 1.7 × 10−8 86.5 2.4 × 10−8 74.8 1.9 × 10−8 76 1.7 × 10−8 64 2.2 × 10−8 52 BTC fit before alkaline treatment BTC fit after alkaline treatment 0.5 to 1.0 M respectively Their binding capacities are also drastically increased compared to the older generation MSS resin, with DBC10% of over 70 mg/mL at 75 cm/h, compared to ∼40 mg/mL for MSS Mass transfer, on the other hand seems to be slower for 650F and PrismA, with a sharper decrease of DBC at higher velocities A quantitative analysis was performed by plotting DBC over residence time and fitting the data with a pore diffusion model (supporting information Fig S1) The effective pore diffusion coefficient for PrismA was determined as De = 2.0 × 10−8 cm−2 /s, which is approximately times lower than that of MSS (De = 5.9 × 10−8 cm−2 /s) The De for 650F was 1.9 × 10−8 cm−2 /s These values (Table 2) are in good agreement with Pabst et al [5] In essence, it can be stated that the high binding capacity of PrismA and 650F comes at the cost of reduced mass transfer rate, as can be seen by the shallower BTCs in Fig This is most likely due to differences in the particle and pore size of the different resins (Table 1) To explore the alkaline resistance of the three resins, they were incubated with high concentrations of NaOH over extended periods of time Specifically, 0.1 M, 0.5 M and M NaOH was used for MSS, 650F and PrismA, respectively, which correspond to the maximum concentrations recommended by the manufacturers BTCs were determined with IgG before and after alkaline treatment While the DBC10% of MSS decreased to less than half of its initial value, both PrismA and 650F show a slight increase in DBC (Fig 2) Alkaline treatment is expected to result in a decrease of static binding capacity The observed increase in DBC for PrismA and 650F could therefore be a result of increased mass transfer To quantify the effect of alkaline treatment on static binding capacity and mass transfer, PrismA was incubated with high NaOH concentrations (0.85, 0.95 and 1.0 M) for up to 48 h BTCs were performed with IgG on the treated resin and the static binding capacity (qm ) and De were determined, with an emphasis on fitting the early breakthrough well (see supplementary material Fig S2) It has to be noted that the fitted De values on the BTC are not identical to the values determined with the DBC versus residence time fit (see Table 2), but the values are still in the range reported by Pabst et al [5] To explore the influence of time and NaOH concentration on an alkaline resistant resin, PrismA was treated with 0.85, 0.95 and M NaOH for 24, 36 and 48 h A negative effect of alkaline treatment on qm can be seen in Fig 3A, were static binding 2.2.3 Adsorption isotherms All experiments were performed in mL vials at room temperature in triplicates Stock solutions of IgG and a 50% (v/v) resin slurry were prepared 25, 50 or 75 μL of slurry was added to a solution volume of 925, 950 or 975 μL to reach a total volume of 10 0 μL in different vials which were then rotated end-of-end described above The solution was filtered through a 0.22 μm syringe filter (SLGVX13NL, Millex-GV filter, Merck, Darmstadt, Germany) and analyzed by UV-VIS photometry at 280 nm assuming an extinction coefficient of 1.4 for IgG Calculation of bound protein was performed by mass balance Experimental data was fitted by the Langmuir adsorption isotherm to obtain the values for the equilibrium association constant Ka and maximum binding capacity qm The adsorption isotherms were determined with native and alkaline treated resin, as described above 2.2.4 Confocal laser scanning microscopy Pure IgG was labelled following the supplier instructions In brief, IgG was incubated in a pH 8.5 sodium bicarbonate buffer with a dye-to-protein molar ratio of 1:3 for h at room temperature in the dark After reaction, unreacted dye was separated by size exclusion chromatography using a PD-10 DG desalting column Average labeling ratios of 22–24% were obtained Batch CLSM experiments were done by placing resin in vials containing 1.5 mL of each labeled protein diluted with sufficient unlabeled protein to yield a final dye-to-protein ratio of 1:40, and rotated end-over-end on a rotator At periodic time intervals, small samples were removed from the vials and rapidly centrifuged at 13,0 0 rpm for 30 s to separate the resin from the supernatant Ratios of resin to protein were chosen to change the protein concentration in the supernatant by a maximum of 10% Results and discussion General mass transfer properties of MabSelect SuRe (MSS), MabSelect PrismA (PrismA) and Toyopearl AF-rProtein A HC-650F (650F) were assessed by breakthrough curves (BTC) Columns with 1–2 mL bed volume were overloaded with IgG at varying linear velocities (Fig 1) The newer generation resins, 650F and PrismA, are both resistant to high concentrations of NaOH, N Lingg, A Daxbacher, D Womser-Matlschweiger et al Journal of Chromatography A 1673 (2022) 463058 Fig BTCs with IgG on MabSelect SuRe (left panel), MabSelect PrismA (middle panel), Toyopearl AF-rProtein A HC-650F (right panel) after varying incubation time with M NaOH Column height was ∼10 cm, velocity was 100 cm/h Fig Time and concentration dependent effect of alkaline treatment on MabSelect PrismA Incubation with 0.85 M, 0.9 M and 1,0 M NaOH was performed for 24, 36 and 48 h, respectively qm (mg IgG adsorbed per mL bed volume) and De (cm2 /s) were obtained from fitting the experimental profiles with a pore diffusion model as shown in Fig S2 capacity is decreased by ∼15% after incubation with 0.95 M NaOH for 48 h The apparent De increases with caustic concentration and prolonged incubation time This suggests that mass transfer of the resin is increasing Adsorption isotherms were determined to investigate if binding affinities were influenced by alkaline treatment for the resins that showed higher alkaline stability in Fig Based on the results shown in Fig for PrismA, alkaline treatment was performed with 1.0 M NaOH for 36 h For 650F exposure was arbitrarily reduced to 0.5 M NaOH for 10 h, respectively, due to the lower stability indicated by the supplier Both native resins show typical rectangular Langmuir adsorption behavior, as expected for affinity chromatography After alkaline treatment, both resins exhibited a decreased qm and Ka , i.e both the binding affinity and maximum capacity were affected negatively (Fig 4) Nonetheless, the measured affinities can still be considered as very high In Fig 4, results are shown for mAb Similar results were obtained with polyclonal IgG (supporting information Fig S3) Such a decrease in affinity after alkaline treatment was also shown by Yang et al for MabSelect ligand with surface plasmon resonance [19] The kinetics of IgG binding to Protein A are still extremely fast, so no kinetic binding effects can be expected To obtain further insight into changes of intraparticle transport, intraparticle antibody uptake curves were measured by CLSM This was only possible for the agarose based MSS and PrismA resins Due to the opacity of the methacrylate backbone of 650F, CLSM measurements are not possible under the same conditions as the agarose-based resins The comparison of uptake profiles for the IgG is shown in Figs 5A and 6A for MSS and PrismA respectively Both resins exhibit typical sharp profiles with a shrinking core which is typical for pore diffusion and highly favorable isotherm Faster mass transfer of MSS is clearly visible despite the smaller particle diameter of PrismA In Figs 5C and 6C the uptake of the IgG on MSS and PrismA, which had been exposed to 0.5 M or 1.0 M NaOH for 44 h are shown The profiles not show sharp fronts but progressively become more diffuse (smoother) and eventually reach the core of the particles much faster The intensity of the saturated particle is reduced, suggesting lower saturation and thus capacity since both channels were recorded with the same laser settings The very diffuse adsorption front is indicative for two situations: (1) pore diffusion with less favorable isotherm or (2) partial or complete surface diffusion Based on the adsorption isotherm data (Fig 4) the latter is the more likely mechanism Artificial CLSM images were used to uncover the underlying transport mechanisms observed in Figs and The artificial images were created based on previously developed methodology by Beck et al [23] Pore and solid diffusion parameters were selected to match the overall observed De values from BTC experiments and to match the experimental CLSM images As shown in Figs 5B and 6B, a good match between artificial and experimental adsorption fronts can be achieved for untreated MSS and PrismA by using a pore-diffusion-controlled mass transfer regime Notably, adsorption fronts in PrismA at 60 and 120 become more diffuse than expected for pore diffusion-controlled mass transfer, which is generally dominated by a shrinking core behavior When simulating the adsorption profiles for alkaline treated resin, the diffuse protein front cannot be achieved by pore diffusion alone (Figs 5D and 6D), especially when using the high Ka values from experimental adsorption isotherms (see Fig 4) To simulate the adsorption profiles for alkaline treated resin, parallel diffusion with a reduced pore diffusion coefficient and a considerable solid diffusion term is able to match the experimental results (Figs 5F and 6F), while parallel diffusion with unchanged pore diffusion terms cannot (Figs 5E and 6E) For alkaline treated PrismA, the artificial CLSM images with parallel diffusion are again a good match for the first 30 min, but adsorption profiles for 60- and 120 deviate We hypothesize that this slow mass transfer in the center of the pore could be due to increasing steric hindrance upon adsorption and reduced accessibility Prior research by Pabst et al suggests that PrismA has a more restrictive pore network as evidenced by the relatively low N Lingg, A Daxbacher, D Womser-Matlschweiger et al Journal of Chromatography A 1673 (2022) 463058 Fig Adsorption isotherms of monoclonal IgG on top MabSelect PrismA and bottom Toyopearl AF-rProtein A HC-650F left before and right after alkaline treatment for 36 and 10 h, respectively The value q refers to adsorbed IgG per mL resin volume Fig CLSM of MabSelect SuRe before and after alkaline treatment A and C show experimental CLSM pictures, whereas B, D, E and F represent artificial CLSM images The observed effective pore diffusion coefficient De from BTCs is given for the experimental CLSM pictures For the artificial images, the pore diffusion coefficient Dp and the solid diffusion coefficient Ds of the underlying model are given N Lingg, A Daxbacher, D Womser-Matlschweiger et al Journal of Chromatography A 1673 (2022) 463058 Fig CLSM of MabSelect PrismA before and after alkaline treatment A and C show experimental CLSM pictures, whereas B, D, E and F represent artificial CLSM images The observed effective pore diffusion coefficient De from BTCs is given for the experimental CLSM pictures For the artificial images, the pore diffusion coefficient Dp and the solid diffusion coefficient Ds of the underlying model are given Fig Linear pH gradient elution of mAb on MabSelect PrismA (left panel) and Toyopearl AF-rProtein A HC-650F (right panel) after varying alkaline treatment εp value for mAb of 0.5 compared to 0.94 for NaCl [5] MSS on the other hand has a much higher ε p of 0.7 for mAb for PrismA and 650F, respectively (Fig 7) A shift towards later retention volume, by approximately column volume, was observed for alkaline treated resin This suggests that the desorption behavior of the ligand is also affected by the alkaline treatment We also investigated if a change in porosity could be responsible for the observed differences A salt pulse performed before and after alkaline treatment (Fig S4) did not indicate any changes in intraparticle porosity However, this retention time shift seems to be contrary to the adsorption behavior as lower pH is required for elution indicating high binding strength at acidic conditions One possible explanation might be deamidation of glutamine and asparagine residues into glutamate and aspartate on the Protein A ligand [25] Binding at neutral pH would then be affected by the newly created negative charges in the Protein A ligand, resulting in the observed decrease in binding affinity Furthermore, a deamidated ligand would require a lower pH to be fully protonated and The emergence of significant solid diffusion after alkaline treatment might be related to the increase of the desorption rate koff in alkaline treated Protein A ligand, that was observed by Yang et al in SPR assays [19] They observed that the decrease in affinity constant Ka was driven by this increase in koff and not a decrease in adsorption rate kon Our own data (Fig and supporting information Fig S3) confirm a slightly reduced, but still very high affinity of IgG for the ligand This results in high partitioning of IgG into the solid phase, but higher mobility due to increased koff , enabling surface diffusion The effect of alkaline treatment on elution behavior of antibodies was investigated on PrismA and 650F resins Elution of IgG with a linear pH gradient was performed at the start and the end of the treatment but also after intermediate time points of 24 h and h N Lingg, A Daxbacher, D Womser-Matlschweiger et al Journal of Chromatography A 1673 (2022) 463058 thus uncharged, to elicit elution Further studies on a molecular level would be required to confirm this hypothesis It has to be noted that the alkaline treatment used in this study differs from regular industrial procedure Instead of using repeated 30 exposure cycles, we have simulated the repeated CIP stress by a single prolonged incubation A final consideration is related to a change of pore structure An investigation of the beads with SEM proofed difficult and showed inconclusive results Obtaining reproducible micrographs that allow to compare the pore structure of even the same bead is challenging Accordingly, no change in pore structure or diameter between virgin and alkaline treated material was observed (supporting information Fig S5) A TEM inspection such as those performed by Zang et al [17,18] and Pathak and Rathore [16] could be informative in this respect but was not available for the present study [4] Z Liu, S.S Mostafa, A.A Shukla, A comparison of Protein A chromatographic stationary phases: performance characteristics for monoclonal antibody purification, Biotechnol Appl Biochem 62 (1) (2015) 37–47, doi:10.1002/bab.1243 [5] T.M Pabst, J Thai, A.K Hunter, Evaluation of recent Protein A stationary phase innovations for capture of biotherapeutics, J Chromatogr A 1554 (2018) 45–60, doi:10.1016/j.chroma.2018.03.060 [6] M Grom, M Kozorog, S Caserman, A Pohar, B Likozar, Protein A affinity chromatography of Chinese hamster ovary (CHO) cell culture broths containing biopharmaceutical monoclonal antibody (mAb): experiments and mechanistic transport, binding and equilibrium modeling, J Chromatogr B Analyt Technol Biomed Life Sci 1083 (2018) 44–56, doi:10.1016/j.jchromb.2018.02.032 [7] R Hahn, K Shimahara, F Steindl, A Jungbauer, Comparison of Protein A affinity sorbents III Life time study, J Chromatogr A 1102 (1–2) (2006) 224–231, doi:10.1016/j.chroma.2005.10.083 [8] M Zarrineh, I.S Mashhadi, M Farhadpour, A Ghassempour, Mechanism of antibodies purification by Protein A, Anal Biochem 609 (2020) 113909, doi:10 1016/j.ab.2020.113909 [9] S Hober, K Nord, M Linhult, Protein A chromatography for antibody purification, J Chromatogr B Analyt Technol Biomed Life Sci 848 (1) (2007) 40–47, doi:10.1016/j.jchromb.2006.09.030 [10] A Gronberg, M Eriksson, M Ersoy, H.J Johansson, A tool for increasing the lifetime of chromatography resins, MAbs (2) (2011) 192–202, doi:10.4161/ mabs.3.2.14874 [11] B Nilsson, T Moks, B Jansson, L Abrahmsen, A Elmblad, E Holmgren, C Henrichson, T.A Jones, M Uhlen, A synthetic IgG-binding domain based on staphylococcal Protein A, Protein Eng (2) (1987) 107–113, doi:10.1093/protein/1.2 107 [12] R Li, V Dowd, D.J Stewart, S.J Burton, C.R Lowe, Design, synthesis, and application of a Protein A mimetic, Nat Biotechnol 16 (2) (1998) 190–195, doi:10.1038/nbt0298-190 [13] M Linhult, S Gulich, T Graslund, A Simon, M Karlsson, A Sjoberg, K Nord, S Hober, Improving the tolerance of a Protein A analogue to repeated alkaline exposures using a bypass mutagenesis approach, Proteins 55 (2) (2004) 407– 416, doi:10.1002/prot.10616 [14] K Minakuchi, D Murata, Y Okubo, Y Nakano, S Yoshida, Remarkable alkaline stability of an engineered Protein A as immunoglobulin affinity ligand: c domain having only one amino acid substitution, Protein Sci 22 (9) (2013) 1230–1238, doi:10.1002/pro.2310 [15] E Muller, J Vajda, Routes to improve binding capacities of affinity resins demonstrated for Protein A chromatography, J Chromatogr B Analyt Technol Biomed Life Sci 1021 (2016) 159–168, doi:10.1016/j.jchromb.2016.01.036 [16] M Pathak, A.S Rathore, Mechanistic understanding of fouling of Protein A chromatography resin, J Chromatogr A 1459 (2016) 78–88, doi:10.1016/j chroma.2016.06.084 [17] S Zhang, W Daniels, J Salm, J Glynn, J Martin, C Gallo, R Godavarti, G Carta, Nature of foulants and fouling mechanism in the Protein A MabSelect resin cycled in a monoclonal antibody purification process, Biotechnol Bioeng 113 (1) (2016) 141–149, doi:10.1002/bit.25706 [18] S Zhang, K Xu, W Daniels, J Salm, J Glynn, J Martin, C Gallo, R Godavarti, G Carta, Structural and functional characteristics of virgin and fouled Protein A MabSelect resin cycled in a monoclonal antibody purification process, Biotechnol Bioeng 113 (2) (2016) 367–375, doi:10.1002/bit.25708 [19] L Yang, J.D Harding, A.V Ivanov, N Ramasubramanyan, D.D Dong, Effect of cleaning agents and additives on Protein A ligand degradation and chromatography performance, J Chromatogr A 1385 (2015) 63–68, doi:10.1016/j.chroma 2015.01.068 [20] C Jiang, J Liu, M Rubacha, A.A Shukla, A mechanistic study of Protein A chromatography resin lifetime, J Chromatogr A 1216 (31) (2009) 5849–5855, doi:10.1016/j.chroma.2009.06.013 [21] Tosoh, TOYOPEARL AF-rProtein A-650F product description, 2021 https: //www.separations.eu.tosohbioscience.com/solutions/process- media- products/ by-mode/protein-a/toyopearl-af-rprotein-a-650f (Accessed 2021-12-10) [22] Cytiva, Benchmarking study: alkaline stability of modern Protein A chromatography resins, 2021 https://www.cytivalifesciences.com/en/us/solutions/ bioprocessing/knowledge- center/alkaline- stability- of- protein- a- affinity- resins (Accessed 2021-12-10) [23] J Beck, E von Lieres, N Zaghi, S Leweke, G Carta, R Hahn, Patterns of protein adsorption in ion-exchange particles and columns: evolution of protein concentration profiles during load, hold, and wash steps predicted for pore and solid diffusion mechanisms, J Chromatogr A 1653 (2021) 462412, doi:10.1016/j.chroma.2021.462412 [24] T.W Weber, R.K Chakravorti, Pore and Solid diffusion models for fixed-bed adsorbers, Aiche J 20 (2) (1974) 228–238 DOI, doi:10.10 02/aic.69020 0204 [25] M Wetterhall, M Ander, T Bjorkman, S Musunuri, R Palmgren, G Rodrigo, Investigation of alkaline effects on Protein A affinity ligands and resins using high resolution mass spectrometry, J Chromatogr B Analyt Technol Biomed Life Sci 1162 (2021) 122473, doi:10.1016/j.jchromb.2020.122473 Conclusion Breakthrough curves of next generation Protein A resins become steeper after alkaline treatment, eventually leading to increased dynamic binding capacity This stems from an increase of mass transfer, driven by a change from pore to parallel diffusion Slight ligand degradation leads to a contribution of solid diffusion as shown by analyzing the intraparticle adsorption profiles Alkaline treatment does not appear to change the pore structure of the bead The ligand degradation also impacts elution behavior, leading to a lower pH for complete elution This behavior can be critical after prolonged resin cycling, when full recovery cannot be reached with standard elution conditions Finally, our findings can help explain anomalous mass transfer and elution effects observed in large scale production with cycled alkaline resistant resins Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper CRediT authorship contribution statement Nico Lingg: Conceptualization, Writing – original draft, Writing – review & editing Andreas Daxbacher: Investigation, Methodology Desiree Womser-Matlschweiger: Investigation, Methodology Dietmar Pum: Investigation, Methodology Jürgen Beck: Formal analysis, Methodology, Visualization Rainer Hahn: Conceptualization, Methodology, Resources, Writing – original draft, Writing – review & editing, Supervision, Visualization Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.chroma.2022.463058 References [1] G.R Bolton, K.K Mehta, The role of more than 40 years of improvement in Protein A chromatography in the growth of the therapeutic antibody industry, Biotechnol Prog 32 (5) (2016) 1193–1202, doi:10.1002/btpr.2324 [2] R Hahn, R Schlegel, A Jungbauer, Comparison of Protein A affinity sorbents, J Chromatogr B Analyt Technol Biomed Life Sci 790 (1–2) (2003) 35–51, doi:10.1016/s1570-0232(03)0 092-8 [3] R Hahn, P Bauerhansl, K Shimahara, C Wizniewski, A Tscheliessnig, A Jungbauer, Comparison of Protein A affinity sorbents II Mass transfer properties, J Chromatogr A 1093 (1–2) (2005) 98–110, doi:10.1016/j.chroma.2005.07.050 ... new resin having a shallower breakthrough than alkaline treated resin Jiang et al investigated the in? ??uence of alkaline CIP cycles on rProtein A Sepharose Fast Flow and MabSelect resins using pure... resins are not alkaline stable and as such 50–100 mM NaOH was used for cleaning Adsorption isotherms and breakthrough curves showed a reduction in static and dynamic binding capacity, and a change... highly alkaline stable Protein A resin MabSelect PrismA and its predecessor MabSelect SuRe To broaden the scope of investigation, a methacrylate-based resin, Toyopearl AF-rProtein A HC-650F, was also